Purpose: Heat shock protein-90 (HSP-90), a molecular chaperone required by numerous oncogenic kinases [e.g., HER-2, epidermal growth factor receptor (EGFR), Raf-1, v-Src, and AKT] for conformational stability, has attracted wide interest as a novel target for cancer therapy. HSP-90 inhibition induces degradation of HSP-90 client proteins, leading to a combinatorial inhibition of multiple oncogenic signaling pathways with consecutive growth arrest and apoptosis. MET, a tyrosine kinase that is constitutively active in tumor cells with MET oncogene amplification, has recently been identified as another HSP-90 client.

Experimental Design: The aim of our study was to assess the efficacy of SNX-2112, a synthetic HSP-90 inhibitor, in 3 different MET-amplified tumor cell lines (GTL-16, MKN-45, and EBC-1) as well as PR-GTL-16 cells, a GTL-16 subline selected for resistance to the highly selective MET kinase inhibitor PHA-665752.

Results: In all cell lines, SNX-2112 led to degradation of MET, HER-2, EGFR, and AKT, as well as abrogation of Ras/Raf/MEK/MAPK and PI3K/AKT signaling, followed by complete cell cycle arrest. SNX-5542, an orally bioavailable prodrug of SNX-2112, displayed significant antitumor efficacy in vivo in nude mice bearing MET-amplified tumor xenografts. Importantly, HSP-90 inhibition maintained its antitumor efficacy in PR-GTL-16 cells both in vitro and in vivo, suggesting that HSP-90 inhibition could be a particularly valuable strategy in MET-amplified tumors that have acquired resistance to MET kinase inhibition.

Conclusions: Our study provides evidence for the efficacy of HSP-90 inhibition in MET-amplified cancer cells, particularly when MET kinase inhibitor resistance has emerged. Clin Cancer Res; 17(1); 122–33. ©2011 AACR.

Translational Relevance

Amplification of the MET oncogene is observed in approximately 15% of gastric cancers and has recently attracted wide interest as a mechanism of acquired resistance to selective epidermal growth factor receptor (EGFR) inhibition in EGFR-mutant non–small-cell lung cancer. Several selective small molecule MET inhibitors have been developed but these have so far failed to show significant in vivo activity in patients, suggesting a need for therapeutic alternatives. In this study, we show that inhibition of heat shock protein-90 (HSP-90) has strong efficacy against MET-amplified tumor cells both in vitro and in vivo. Of particular importance, HSP-90 inhibition maintains its in vitro and in vivo antitumor efficacy even in MET-amplified tumor cells that have become resistant to highly selective MET kinase inhibition. We conclude that HSP-90 inhibition is a promising therapeutic strategy forMET-amplified tumors, particularly if secondary resistance against selective MET inhibition has emerged.

The MET proto-oncogene encodes a heterodimeric receptor tyrosine kinase that consists of an extracellular α-chain and a transmembrane β-chain (1, 2). Hepatocyte growth factor binds to MET with high affinity and induces receptor dimerization with consecutive triggering of MET tyrosine kinase activity (3, 4). MET downstream signaling includes the phosphoinositide-3-kinase (PI3K)/AKT, Ras/Raf/ MAP/ERK kinase (MEK)/mitogen-activated protein kinase (MAPK), and phospholipase C-γ pathways (3).

In tumors in which MET is overexpressed due to genomic amplification, MET becomes constitutively active, rendering the malignant cells highly dependent on MET signaling for proliferation and survival (“MET oncogene addiction”; refs. 5, 6). Several selective small molecule MET inhibitors have been developed and have shown promising antitumor efficacy against MET-amplified tumor cells both in vitro as well as in mouse models (7–9). However, as with other highly selective kinase inhibitors, acquired resistance may develop, which prompted us to investigate possible therapeutic alternatives.

A large number of oncogenic kinases [e.g., HER-2, epidermal growth factor receptor (EGFR), v-Src, Raf1, cyclin-dependent kinase-4, and AKT] require heat shock protein-90 (HSP-90) for conformational stability (10–17). Given the critical roles played by these HSP-90 clients in tumor cell signaling, proliferation, and survival, inhibition of HSP-90 has emerged as a potent antitumor treatment strategy (18). The underlying mechanism involves proteasomal degradation of HSP-90 client proteins leading to disruption of the tumor cell signaling network with consecutive cell cycle arrest and apoptosis.

Previous studies have shown that multiple HSP-90 clients are activated in MET oncogene-addicted cancer cells either through MET-dependent downstream signaling or receptor cross talk (e.g., Raf1, AKT, and EGFR; refs. 5–7, 19). Furthermore, MET itself has recently been implicated as an HSP-90 client (20–24). We therefore hypothesized that HSP-90 inhibition could be a particularly promising treatment strategy in MET-amplified cancer cells. Moreover, due to its combined effect on multiple signal transduction pathways, we hypothesized that HSP-90 inhibition could also overcome acquired resistance to small molecule MET inhibition in these malignancies. In this study, we have tested the effects of SNX-2112, a novel synthetic HSP-90 inhibitor (25–27), in 3 different tumor cell lines with MET amplification [EBC-1 (non–small-cell lung cancer), GTL-16 (gastric cancer), and MKN-45 (gastric cancer)] as well as PHA-resistant (PR)-GTL-16 cells that we selected for acquired resistance to PHA-665752, a highly selective MET kinase inhibitor. In all cells, degradation of MET was observed together with degradation of the HSP-90 clients HER-2, EGFR, and AKT. MET degradation was paralleled by loss of MET phosphorylation, abrogation of downstream PI3K/AKT, and Ras/Raf/MEK/MAPK signaling as well as by cell cycle arrest. HSP-90 inhibition using SNX-5542, an orally bioavailable prodrug of SNX-2112, also displayed significant antitumor activity in vivo in nude mice bearing MET-amplified xenografts with minimal toxicity. Importantly, HSP-90 inhibition maintained its in vitro and in vivo antitumor efficacy in PR-GTL-16 tumor cells with acquired resistance to PHA-665752, providing a strong rationale for the use of HSP-90 inhibition in MET-amplified tumors that have become resistant to selective MET kinase inhibition.

Cell lines

Human GTL-16 gastric cancer cells were a gift from Dr. Silvia Giordano (Institute for Cancer Research and Treatment, Torino School of Medicine, Italy). MKN-45 gastric cancer cells were obtained from the RIKEN BRC Cell Bank (RIKEN BioResource Center). EBC-1 non–small-cell lung cancer cells were from the Health Science Research Resources Bank (Japan Health Sciences Foundation). NCI-H820 cells were obtained from the American Type Culture Collection. GTL-16 and PR-GTL-16 cells were grown in Dulbecco's modified Eagle's Medium, MKN-45 and NCI-H820 cells were grown in RPMI-1640, and EBC-1 cells were grown in Eagle's minimal essential medium + 2 mmol/L l-glutamine + 1 mmol/L sodium pyruvate + 0.1 mmol/L nonessential amino acids. All media were supplemented with 10% fetal calf serum and maintained at 37°C in a humidified atmosphere containing 5% CO2.

Chemicals

PHA-665752, PD173074, and PD330631 were provided by Pfizer Global Research and Development. Gefitinib (ZD-1839; Iressa) was obtained from AstraZeneca Pharmaceuticals. Recombinant human fibroblast growth factor-3 (FGF-3) was purchased from R&D Systems. SNX-2112 (for chemical structure see Supplementary Fig. S1) and SNX-5542 were obtained from Serenex, Inc. SNX-2112 was dissolved in dimethyl sulfoxide (DMSO) for in vitro studies, whereas SNX-5542 was formulated in 5% dextrose in water for in vivo studies.

Western blot

After removal of growth medium, tissue culture flasks were placed on ice and washed twice with ice-cold Tris-buffered saline (TBS). Cells were scraped off the culture flasks, centrifuged, and placed in ice-cold cell lysis buffer (Cell Signaling Technology) containing protease and phosphatase inhibitors (Halt protease and phosphatase inhibitor cocktails; Pierce). After shaking for 15 minutes at 4°C, the lysates were centrifuged at 20,000 × g for 15 minutes and stored at −70°C until further use. For Western blotting, equal amounts of protein (50 μg) were boiled in Laemmli buffer for 5 minutes, resolved by 10% SDS-polyacrylamide gel electrophoresis (Invitrogen) and electrophoretically transferred onto a polyvinylidene difluoride membrane (BioRad). After blocking nonspecific binding sites with 5% nonfat dry milk in TBS + 0.05% Tween 20 (TBS-T), the membrane was incubated with the respective primary antibodies. After 3 washes with TBS-T, the membrane was incubated for 1 hour at room temperature with a horseradish peroxidase-linked secondary antibody, followed by several washes with TBS-T. The immunocomplexes were visualized using the ECL Plus detection system (GE Healthcare).

Antibodies

Antibodies against MET (C-12), EGFR (1005), and cyclin D1 (M-20) were from Santa Cruz Biotechnology. Antibodies against phospho-MET (Tyr1230/1234/1235), p44/42 MAPK, phospho-p44/42 MAPK (Thr202/Tyr204), AKT, and phospho-AKT (Ser473) were from Cell Signaling Technology. Anti–HER-2 antibody (clone e2-4001 + 3B5) was from Lab Vision.

Cell proliferation assays

Cellular proliferation was measured using a commercially available 5-bromo-2-deoxyuridine (BrdU) cell proliferation assay (Roche). Briefly, the cells were seeded in triplicate in flat-bottom 96-well plates at 5,000 cells per well and allowed to adhere for 48 hours. Thereafter, the cells were treated for 24 hours, as indicated. After incubation with BrdU labeling reagent for 2 hours, the cells were fixed and BrdU incorporation into newly synthesized DNA was assessed by incubation with an anti-BrdU peroxidase-conjugated antibody for 90 minutes, followed by addition of substrate solution and colorimetric detection at 450 and 690 nm, respectively. IC50 values were calculated by 4-parameter curve fitting using SigmaPlot 11.0 software.

Cell cycle analysis

Cells were seeded in 100-mm dishes at a density of 5 × 105 cells per dish. Forty-eight hours later, the cells were treated with drug or vehicle (DMSO) for 24 hours. Both adherent and floating cells were harvested and stained with ethidium bromide. Quantitation of the cell cycle distribution was performed by flow cytometric analysis.

Apoptosis

Induction of apoptosis was assessed using the Annexin V FITC Apoptosis Detection Kit from BD Pharmingen (BD Biosciences). Briefly, the cells were treated with SNX-2112 for 24 hours, stained with Annexin V FITC + propidium iodide (PI) at room temperature according to the manufacturer's instructions, and finally subjected to flow cytometric analysis.

Creation of PHA-665752–resistant GTL-16 cells in vitro

To generate GTL-16 cells resistant to the MET tyrosine kinase inhibitor PHA-665752, cells were exposed to increasing concentrations of PHA-665752 in vitro. After 6 months, a resistant subclone (PR-GTL-16) was obtained, which grew in the presence of up to 2 μmol/L PHA-665752 with similar growth kinetics as the untreated GTL-16 cells. After several passages in the absence of PHA-665752, PR-GTL-16 cells retained their resistant phenotype as determined by repeated cell proliferation assays. Parental GTL-16 cells were maintained concomitantly without PHA-665752 and no significant change in the sensitivity to PHA-665752 was noted.

RNA isolation, labeling, and microarray analysis

Total RNA was isolated from tissue using the RNeasy Kit (Qiagen). The quality of the RNA was ensured before labeling by analyzing 20 to 50 ng of each sample using the RNA 6000 NanoAssay and the Bioanalyzer 2100 (Agilent Technologies). Samples with a 28S/18S ribosomal peak ratio of 1.8 to 2.0 and a RIN number >7.0 were considered suitable for labeling. For samples meeting this standard, 2 μg of total RNA was used for cDNA synthesis using an oligo-dT-T7 primer and the SuperScript Double-Stranded cDNA Synthesis Kit (Invitrogen). Synthesis, linear amplification, and labeling of cRNA were accomplished by in vitro transcription using the MessageAmp aRNA Kit (Ambion) and biotinylated nucleotides (Enzo Diagnostics). Ten micrograms of labeled and fragmented cRNA were then hybridized to the Human HG-U133A2.0 GeneChip (Affymetrix) at 45°C for 16 hours. After washing, the chips were processed according to the manufacturer's instructions and scanned with a high-numerical aperture and flying objective lens in the GS3000 scanner (Affymetrix). The image was quantified using GCOS version 1.4 (GeneChip Operating Software, Affymetrix).

Reverse transcriptase PCR

TaqMan Gene Expression Assays for MET, FGF-3, and 18s ribosomal RNA were purchased from Applied Biosystems. Gene expression was measured using the ABI Prism 7900HT Sequence Detection System from Applied Biosystems. Reverse transcriptase PCR (RT-PCR) of cDNA specimens was conducted as described previously (28).

siRNA

For siRNA experiments, cells were seeded in triplicate in 96-well plates at 5,000 cells per well in antibiotic-free complete medium and were allowed to adhere for 24 hours at 37°C. Thereafter, the cells were transfected with siGenome ON-TARGET plus FGF-3 or Non-Silencing Pool #3 siRNA (Dharmacon) according to the manufacturer's instructions. Untransfected cells were used for control. After 48 hours, cell proliferation was determined using a BrdU cell proliferation assay.

Animal studies

Four- to 6-week-old nu/nu athymic BALB/c mice were obtained from the National Cancer Institute-Frederick Cancer Center and maintained in pressurized ventilated caging at the Sloan-Kettering Institute (New York, NY). All studies were done in compliance with Institutional Animal Care and Use Committee guidelines. Tumors were established by flank injection of 1 × 107 cells suspended with reconstituted basement membrane (Matrigel; Collaborative Research) at a ratio of 1:1 (volume). For efficacy studies, mice (5 per group) with established tumors were selected. Fourteen days after inoculation, mice were treated with SNX-5542, using the indicated doses. Tumor dimensions were measured with vernier calipers and tumor volumes were calculated using the formula: π/6 × larger diameter × (smaller diameter)2.

Statistical analysis of xenograft experiments

Because of right-skewed distribution of tumor volume measurements, a logarithmic transformation of the tumor volume was performed. For data depiction, mean and corresponding 68% confidence intervals (CI) were calculated and transformed back to the original scale (which would correspond to mean ± SE for normally distributed data). Group differences in tumor volume for the different cell lines and effects over time were modeled with a linear mixed model with repeated measures. A logistic transformation was applied to the dependent variable tumor volume to achieve homoscedasticity of residuals. Group differences, effect of time, and a group–time interaction were modeled and tested. All P values are 2-sided and P ≤ 0.05 was considered significant. Calculations were performed using the statistical software package SAS(r) (SAS Institute Inc.; version 9.2).

HSP-90 inhibition induces degradation of MET and other HSP-90 client proteins in tumor cells with MET amplification

We first assessed the effects of the synthetic HSP-90 inhibitor SNX-2112 on MET expression in 3 different MET-amplified tumor cell lines. As shown in Fig. 1A, SNX-2112 induced a dose-dependent degradation of MET in all cell lines, together with degradation of the HSP-90 client proteins HER-2, EGFR, and AKT. Degradation of MET was paralleled by complete dephosphorylation of MET, inhibition of downstream MAPK and AKT activation, as well as loss of cyclin D1 expression, indicating cell cycle arrest.

Figure 1.

HSP-90 inhibition induces degradation of MET and inhibition of proliferation in tumor cells with MET amplification. A, tumor cells were treated with SNX-2112 for 24 hours as indicated, followed by Western blot analysis. B, comparative kinetics of HSP-90 and selective MET kinase inhibition in GTL-16 gastric cancer cells. Cells were either treated with 1,000 nmol/L SNX-2112 or 400 nmol/L PHA-665752 as indicated, followed by Western blot analysis. C, cells were treated for 24 hours as indicated, followed by cell proliferation analysis. Data are depicted as mean ± SE.

Figure 1.

HSP-90 inhibition induces degradation of MET and inhibition of proliferation in tumor cells with MET amplification. A, tumor cells were treated with SNX-2112 for 24 hours as indicated, followed by Western blot analysis. B, comparative kinetics of HSP-90 and selective MET kinase inhibition in GTL-16 gastric cancer cells. Cells were either treated with 1,000 nmol/L SNX-2112 or 400 nmol/L PHA-665752 as indicated, followed by Western blot analysis. C, cells were treated for 24 hours as indicated, followed by cell proliferation analysis. Data are depicted as mean ± SE.

Close modal

SNX-2112–induced dephosphorylation of MET, AKT, and MAPK occurs approximately 6 hours after initiation of treatment, whereas selective MET inhibition using the small molecule inhibitor PHA-665752 causes immediate dephosphorylation of MET and downstream signaling pathways

As shown in Fig. 1B, dephosphorylation of MET, AKT, and MAPK does not occur before 6 hours of SNX-2112 treatment, whereas PHA-665752, a highly selective small molecule inhibitor of MET, causes complete dephosphorylation of these proteins as early as 15 minutes after onset of treatment. However, in contrast to the HSP-90 inhibitor, PHA-665752 does not affect the total levels of MET, AKT, HER-2, and EGFR within the cells. After 24 hours of treatment, complete loss of cyclin D1 expression is observed both in SNX-2112- and PHA-665752-treated cells (Fig. 1B).

Effects of SNX-2112 on cell proliferation of tumor cells with MET amplification

Next, we assessed the effects of SNX-2112 on cell proliferation of MET-amplified cell lines. In all cells, a potent growth-inhibitory effect was observed. IC50 values for the individual cell lines were 25.2 (EBC-1), 30.3 (MKN-45), and 35.6 nmol/L (GTL-16; Fig. 1C).

SNX-2112 induces cell cycle arrest in tumor cells with METamplification

Figure 2 shows the cell cycle data obtained after 24 hours of SNX-2112 treatment: Compatible with the cell proliferation results, SNX-2112 induced G1-arrest at a concentration of 50 nmol/L in all cell lines, whereas at higher concentrations (100 and 1,000 nmol/L) an increasing proportion of the cells accumulated in G2.

Figure 2.

Cell cycle analysis. Cells were treated for 24 hours with increasing doses of SNX-2112, followed by flow cytometric analysis of the cell cycle distribution.

Figure 2.

Cell cycle analysis. Cells were treated for 24 hours with increasing doses of SNX-2112, followed by flow cytometric analysis of the cell cycle distribution.

Close modal

Efficacy of HSP-90 inhibition in a MET-amplified xenograft model

To study the antitumor efficacy of HSP-90 inhibition in MET-amplified tumors in vivo, nude mice bearing xenografts of GTL-16 and EBC-1 tumor cells were treated with increasing doses of SNX-5542, an orally bioavailable prodrug of SNX-2112. As shown in Fig. 3, treatment of mice with 50 mg/kg of SNX-5542 using a Monday–Wednesday–Friday schedule significantly inhibited growth of GTL-16 and EBC-1 xenografts as compared with the control group (P < 0.0001 and P = 0.0003, respectively) and was well tolerated without measurable toxicity. Importantly, increasing the dose of SNX-5542 to 75 mg/kg led to a further significant enhancement of the growth-inhibitory effect with partial regression of GTL-16 xenografts (P < 0.0001). However, consistent with a previous report by Chandarlapaty and colleagues (25), SNX-5542 at the higher dose of 75 mg/kg also induced an average weight loss of 21.3% with 2 of 5 animal deaths by day 24. Consequently, the maximum SNX-5542 dose was reduced to 60 mg/kg for further experiments, albeit no significant enhancement of the growth-inhibitory effect as compared with the 50 mg/kg was noted (P < 0.7115; Fig. 3, bottom graph).

Figure 3.

Treatment with SNX-5542 induces inhibition of tumor growth in GTL-16 and EBC-1 xenograft models. Four to 6-week-old athymic nu/nu BALB/c mice with established GTL-16 and EBC-1 xenografts (5 per group) were randomized to treatment with SNX-5542 or the control group. Mice were treated with SNX-5542, an orally bioavailable prodrug of SNX-2112, by oral gavage on a Monday–Wednesday–Friday schedule. The tumor volume was measured as described in Materials and Methods. Data are depicted as mean ± 68% CI to account for right-skewed data distribution. This would correspond to a depiction as mean ± SE in the case of normal distribution. P values are from the regression model. The control arm of the EBC-1 xenograft model had to be discontinued after day 22 because of excessive tumor growth and ulceration of the tumors.

Figure 3.

Treatment with SNX-5542 induces inhibition of tumor growth in GTL-16 and EBC-1 xenograft models. Four to 6-week-old athymic nu/nu BALB/c mice with established GTL-16 and EBC-1 xenografts (5 per group) were randomized to treatment with SNX-5542 or the control group. Mice were treated with SNX-5542, an orally bioavailable prodrug of SNX-2112, by oral gavage on a Monday–Wednesday–Friday schedule. The tumor volume was measured as described in Materials and Methods. Data are depicted as mean ± 68% CI to account for right-skewed data distribution. This would correspond to a depiction as mean ± SE in the case of normal distribution. P values are from the regression model. The control arm of the EBC-1 xenograft model had to be discontinued after day 22 because of excessive tumor growth and ulceration of the tumors.

Close modal

Creation of MET kinase inhibitor-resistant GTL-16 tumor cells

To assess whether HSP-90 inhibition maintains its antitumor efficacy in MET-amplified tumor cells with acquired resistance to a selective MET kinase inhibitor, we created GTL-16 cells resistant to the highly selective MET kinase inhibitor PHA-665752. Cells were cultured in vitro in increasing concentrations of PHA-665752 over a 6-month period, resulting in a resistant cell line (PR-GTL-16) that displayed uninhibited growth and cell cycle distribution even in the presence of 800 nmol/L PHA-665752 (Fig. 4A and B).

Figure 4.

Characterization of PHA-665752-resistant (PR) GTL-16 tumor cells. A, GTL-16 and PR-GTL-16 tumor cells were treated with PHA-665752 as indicated. After 24 hours, a BrdU cell proliferation assay was performed. B, cells were treated with 400 nmol/L PHA-665752 for 24 hours. Thereafter, flow cytometric analysis of the cell cycle distribution was performed. C, RT-PCR analysis for 18s, MET, and FGF-3 RNA of PR-GTL-16 as compared with parental GTL-16 tumor cells. CT, cycle threshold. *, in parental GTL-16 cells, no CT value could be determined, indicating that FGF-3 is not expressed in these cells. Data are depicted as mean ± SE.

Figure 4.

Characterization of PHA-665752-resistant (PR) GTL-16 tumor cells. A, GTL-16 and PR-GTL-16 tumor cells were treated with PHA-665752 as indicated. After 24 hours, a BrdU cell proliferation assay was performed. B, cells were treated with 400 nmol/L PHA-665752 for 24 hours. Thereafter, flow cytometric analysis of the cell cycle distribution was performed. C, RT-PCR analysis for 18s, MET, and FGF-3 RNA of PR-GTL-16 as compared with parental GTL-16 tumor cells. CT, cycle threshold. *, in parental GTL-16 cells, no CT value could be determined, indicating that FGF-3 is not expressed in these cells. Data are depicted as mean ± SE.

Close modal

PR-GTL-16 cells display acquired expression of fibroblast growth factor-3 that is not detected in parental GTL-16 cells

Microarray analysis of PR-GTL-16 cells using the HG-U133A2.0 GeneChip revealed that 170 genes were differentially regulated with a fold-change cutoff of ±2 and a P-value cutoff of P < 0.005. Thirty-six of these genes were regulated at least 4-fold. The most highly regulated gene was FGF-3, which displayed a 64-fold increase of expression in PR-GTL-16 as compared with parental GTL-16 cells (Supplementary Table S1), making it the most likely candidate for the resistant phenotype of PR-GTL-16 cells although other changes (e.g., >4-fold upregulation of cyclin D2) should not be remain unmentioned. The microarray data were confirmed by RT-PCR analysis, in which FGF-3 mRNA was readily detectable in PR-GTL-16 cells, whereas it was below the detection limit in parental GTL-16 cells. Importantly, MET mRNA expression did not differ significantly between the cells, indicating that the resistance mechanism of PR-GTL-16 cells does not involve a change of MET expression (Fig. 4C).

FGF-3 confers resistance to MET inhibition through reactivation of AKT and MAPK signaling

Although parental GTL-16 cells displayed complete dephosphorylation of MET, AKT, and MAPK following PHA-665752 treatment, MET but not AKT and MAPK were dephosphorylated in PR-GTL-16 cells (Fig. 5A). On the basis of the microarray and RT-PCR results, we hypothesized that autocrine FGF-3 signaling could be responsible for sustained AKT and MAPK activation in PR-GTL-16 cells. This was supported by the observation that exogenous FGF-3 led to rephosphorylation of AKT and MAPK in parental GTL-16 cells and was able to rescue them from PHA-665752-mediated growth inhibition (Fig. 5A and B, lanes 5). Furthermore, siRNA-mediated knockdown of FGF-3 in combination with PHA-665752 (but neither treatment alone) significantly inhibited proliferation of PR-GTL-16 cells (Fig. 5C) and induced complete dephosphorylation of AKT and MAPK (data not shown). Finally, combined blockade of MET and FGF receptor (FGFR) signaling using PHA-665752 together with either of 2 selective FGFR inhibitors (PD173074 or PD330361) led to a dose-dependent inhibition of PR-GTL-16 proliferation, further illustrating FGF-3/FGFR-mediated signaling as the central mechanism of resistance to MET kinase inhibition in PR-GTL-16 cells (Fig. 5D). To assess in more detail which FGFRs mediate the FGF-3 effect in PR-GTL-16 cells, the expression of FGFR-1 and -2 (the receptors with the highest affinity for FGF-3; ref. 29) was determined using quantitative RT-PCR. Importantly, PR-GTL-16 cells were found to express both FGFR-1 and -2 mRNA, which strongly suggests that both receptors contribute to the FGF-3 rescue effect in these cells. This was further confirmed by FGFR-1/2 gene silencing using shRNA constructs (data not shown).

Figure 5.

FGF-3 stimulates AKT and MAPK phosphorylation and confers resistance to PHA-665752–mediated growth inhibition. A, GTL-16 and PR-GTL-16 cells were treated for 24 hours with 400 nmol/L PHA-665752 with or without addition of 0.1 μmol/L recombinant FGF-3, followed by Western blot analysis. B, cells were treated as in A, followed by a BrdU cell proliferation assay. Data are depicted as mean ± SE. C, PR-GTL-16 cells were transfected with FGF-3 siRNA and concomitantly treated with 400 nmol/L PHA-665752 for 48 hours. Thereafter, cell proliferation was measured using a BrdU cell proliferation assay. For control, cells were transfected with nonsilencing siRNA or left untransfected. Data are depicted as mean ± SE. D, PR-GTL-16 cells were treated with increasing doses of the selective FGF receptor (FGFR) inhibitor PD173074, alone or in combination with 400 nmol/L PHA-665752. In a second set of experiments, PD173074 was replaced by PD330361, another selective small molecule FGFR inhibitor. After 24 hours, cell proliferation was measured using a BrdU cell proliferation assay. Open squares, FGFR inhibitor alone; filled squares, FGFR inhibitor in combination with PHA-665752. Data are depicted as mean ± SE.

Figure 5.

FGF-3 stimulates AKT and MAPK phosphorylation and confers resistance to PHA-665752–mediated growth inhibition. A, GTL-16 and PR-GTL-16 cells were treated for 24 hours with 400 nmol/L PHA-665752 with or without addition of 0.1 μmol/L recombinant FGF-3, followed by Western blot analysis. B, cells were treated as in A, followed by a BrdU cell proliferation assay. Data are depicted as mean ± SE. C, PR-GTL-16 cells were transfected with FGF-3 siRNA and concomitantly treated with 400 nmol/L PHA-665752 for 48 hours. Thereafter, cell proliferation was measured using a BrdU cell proliferation assay. For control, cells were transfected with nonsilencing siRNA or left untransfected. Data are depicted as mean ± SE. D, PR-GTL-16 cells were treated with increasing doses of the selective FGF receptor (FGFR) inhibitor PD173074, alone or in combination with 400 nmol/L PHA-665752. In a second set of experiments, PD173074 was replaced by PD330361, another selective small molecule FGFR inhibitor. After 24 hours, cell proliferation was measured using a BrdU cell proliferation assay. Open squares, FGFR inhibitor alone; filled squares, FGFR inhibitor in combination with PHA-665752. Data are depicted as mean ± SE.

Close modal

HSP-90 inhibition maintains its antitumor efficacy in MET-amplified tumor cells with acquired resistance to a MET kinase inhibitor

Having established a MET-amplified tumor cell line with acquired resistance to MET kinase inhibition, we next sought to determine whether inhibition of HSP-90 maintains its antitumor efficacy in the resistant cells. Figure 6A illustrates that treatment of PR-GTL-16 cells with SNX-2112 led to degradation of MET, AKT, EGFR, and HER-2 as in parental GTL-16 cells. SNX-2112 also induced dephosphorylation of MET, AKT, and MAPK together with loss of cyclin D1 expression, cell cycle arrest, and induction of apoptosis (Fig. 6B and D). The sensitivity of PR-GTL-16 cells toward SNX-2112 was comparable albeit slightly lower than that of parental GTL-16 cells [IC50 = 57.5 nmol/L [PR-GTL-16] versus 35.6 nmol/L (parental GTL-16); Fig. 6B and C]. We also tested the in vivo efficacy of SNX-5542 in nude mice bearing xenografts of PR-GTL-16 cells. As shown in Fig. 6E, SNX-5542 significantly inhibited growth of PR-GTL-16 xenografts, similar to its effect in parental GTL-16 xenografts.

Figure 6.

HSP-90 inhibition maintains its in vitro and in vivo antitumor efficacy in PR-GTL-16 cells. A, PR-GTL-16 cells were treated with SNX-2112 at the indicated doses. After 24 hours, Western blot analysis was performed. B, cells were treated as in A, followed by cell cycle analysis. C, PR-GTL-16 and GTL-16 cells for comparison were treated with increasing doses of SNX-2112 for 24 hours, followed by a BrdU cell proliferation assay. D, GTL-16 and PR-GTL-16 cells were treated with increasing doses of SNX-2112 as indicated, followed by Annexin V/PI staining and flow cytometric analysis. Apoptotic cells: cells staining positive for Annexin V [including early apoptotic cells (Annexin V+/PI−) and late apoptotic cells (Annexin V+/PI+)]. Necrotic cells: cells staining positive for PI without staining for Annexin V (Annexin V−/PI+). Viable cells: cells staining negative for both Annexin V and PI (Annexin V−/PI−). E, nude mice (5 per group) with established PR-GTL-16 xenografts were treated with the indicated doses of SNX-5542 on a Monday–Wednesday–Friday schedule. The tumor volume was measured as described in Materials and Methods. Data are depicted as mean ± 68% CI to account for right-skewed data distribution. This would correspond to a depiction as mean ± SE in the case of normal distribution. P values are from the regression model.

Figure 6.

HSP-90 inhibition maintains its in vitro and in vivo antitumor efficacy in PR-GTL-16 cells. A, PR-GTL-16 cells were treated with SNX-2112 at the indicated doses. After 24 hours, Western blot analysis was performed. B, cells were treated as in A, followed by cell cycle analysis. C, PR-GTL-16 and GTL-16 cells for comparison were treated with increasing doses of SNX-2112 for 24 hours, followed by a BrdU cell proliferation assay. D, GTL-16 and PR-GTL-16 cells were treated with increasing doses of SNX-2112 as indicated, followed by Annexin V/PI staining and flow cytometric analysis. Apoptotic cells: cells staining positive for Annexin V [including early apoptotic cells (Annexin V+/PI−) and late apoptotic cells (Annexin V+/PI+)]. Necrotic cells: cells staining positive for PI without staining for Annexin V (Annexin V−/PI+). Viable cells: cells staining negative for both Annexin V and PI (Annexin V−/PI−). E, nude mice (5 per group) with established PR-GTL-16 xenografts were treated with the indicated doses of SNX-5542 on a Monday–Wednesday–Friday schedule. The tumor volume was measured as described in Materials and Methods. Data are depicted as mean ± 68% CI to account for right-skewed data distribution. This would correspond to a depiction as mean ± SE in the case of normal distribution. P values are from the regression model.

Close modal

Efficacy of HSP-90 inhibition in MET-amplified NCI-H820 cells

Recently, McDermott and colleagues (30) have shown that secondary activation of the EGFR signaling pathway is another mechanism of resistance toward MET kinase inhibition in MET-amplified tumor cells. To assess whether HSP-90 inhibition would also be effective in such MET-amplified tumor cells with secondary activation of EGFR signaling, SNX-2112 was tested in lung cancer-derived NCI-H820 cells that harbor both MET amplification and an activating T790M EGFR mutation. As shown in Supplementary Fig. S2, neither gefitinib (a selective EGFR inhibitor) nor PHA-665752 had an effect on cell proliferation in these cells whereas SNX-2112 led to a >50% inhibition of growth after 24 hours of treatment.

The aim of this study was to assess the effects of HSP-90 inhibition using the novel synthetic HSP-90 inhibitor SNX-2112 in tumor cells with MET amplification. We have found that HSP-90 inhibition induces degradation of MET, inhibition of MET and downstream PI3K/AKT and Ras/Raf/MEK/MAPK signaling as well as cell cycle arrest in 3 different MET-amplified tumor cell lines (EBC-1, GTL-16, and MKN-45). We were also able to show that SNX-5542, an orally bioavailable prodrug of SNX-2112, has significant antitumor efficacy in vivo in nude mice bearing MET-amplified tumor xenografts. Notably, HSP-90 inhibition maintained its in vitro and in vivo efficacy against MET-amplified tumor cells resistant to the highly specific MET kinase inhibitor PHA-665752, suggesting that HSP-90 inhibition could be a particularly valuable strategy in MET-amplified tumors with acquired resistance to selective MET kinase inhibition.

Tumor cells harboring amplification of the MET oncogene are largely dependent on constitutive activation of the MET kinase for proliferation and survival, which renders them highly sensitive toward targeted MET inhibition (6, 7, 19). Several small molecule MET inhibitors have been developed and have recently entered early-stage clinical trials (31, 32). However, as with other highly selective kinase inhibitors (e.g., gefitinib and imatinib), development of acquired resistance is likely. Here, we show that de novo FGF-3 expression is a possible mechanism of resistance to the selective MET kinase inhibitor PHA-665752 in MET-amplified GTL-16 tumor cells. In the resistant cell line (PR-GTL-16), FGF-3, which displays tumorigenic potential in murine breast and colorectal cancer models (33–37), maintains AKT and MAPK phosphorylation even when MET is completely inhibited, whereas knockdown of FGF-3 or selective inhibition of FGFR signaling using either of the 2 selective small molecule FGFR inhibitors (PD173074 or PD330361) largely restores PHA-665752-sensitivity.

Our data add to the growing body of evidence that primary or secondary activation of parallel growth factor receptor signaling resulting in sustained PI3K/AKT and/or Ras/Raf/MEK/MAPK pathway activation is a mechanism of resistance to selective tyrosine kinase inhibition in oncogene-addicted cancer cells. Secondary amplification of MET confers resistance to the EGFR inhibitor gefitinib in EGFR-mutant non–small-cell lung cancer cells (38), loss of insulin-like growth factor (IGF) binding protein-3 expression with consecutive activation of the IGF receptor pathway mediates gefitinib resistance in A431 cells with EGFR amplification (39), and an autocrine FGF-2 signaling loop is associated with resistance to targeted EGFR inhibition in a subset of non–small-cell lung cancer cells (40). Furthermore, it has recently been showed by our group that HER-kinase activation confers resistance to targeted MET inhibition in MET-amplified gastric cancer cells (19).

Because oncogene-addicted tumor cells may readily overcome the effects of highly selective kinase inhibition through activation of redundant survival pathways, solitary inhibition of the oncogenically activated kinase is unlikely to exhibit long-term antitumor efficacy in these malignancies. Instead, the therapeutic strategy should take into account both oncogenic kinase activation as well as potential signaling pathway redundancy, that is, by simultaneously inhibiting multiple key signaling molecules. Conceivably, such an approach would decrease the likelihood of resistance as well as maintain antitumor efficacy in tumors that have become resistant to selective kinase inhibitor treatment.

A large number of oncogenic receptor tyrosine kinases and downstream signaling molecules require HSP-90 for conformational stability, including HER-2, EGFR, v-Src, Raf1, cyclin-dependent kinase-4, and AKT (10–17). Given the important roles played by these HSP-90 clients in signal transduction, proliferation, and survival, inhibition of HSP-90 has emerged as a potential antitumor treatment strategy (18). Its underlying mechanism involves proteasomal degradation of HSP-90 client proteins leading to disruption of the tumor cell signaling network—both at the receptor tyrosine kinase and downstream signaling level. HSP-90 inhibition therefore exemplifies the above-mentioned concept of multitargeted signaling pathway inhibition, making it a highly attractive strategy for the treatment of oncogene-addicted malignancies (41). Indeed, it could be showed that oncogene-addicted tumors (e.g., HER-2-amplified breast cancers or tumors with mutational activation of the EGFR) are particularly sensitive to HSP-90 inhibition both in vitro and in vivo (12, 15, 25).

In MET-amplified cells, a large number of HSP-90 clients (e.g., Raf-1, AKT and EGFR) are activated by MET either through MET-dependent downstream signaling or receptor cross talk (5–7, 19, 38). Moreover, recent studies have indicated that MET itself is an HSP-90 client protein (20–24). We therefore hypothesized that MET oncogene-addicted tumors may be equally amenable to the effects of HSP-90 inhibition. Indeed, our study shows that MET-amplified tumor cells are highly susceptible toward the antitumor effects of HSP-90 inhibition, with a sensitivity that is largely identical to that of HER-2–amplified breast cancer cells.

Of note, HSP-90 inhibition maintained its in vitro and in vivo antitumor efficacy in PR-GTL-16 cells with acquired resistance to the selective MET kinase inhibitor PHA-665752. In these cells, FGF-3 (through its receptors FGFR-1 and -2) induces sustained activation of the PI3K/AKT and Ras/Raf/MEK/MAPK pathways even when MET is completely inhibited, suggesting that combinatorial signal transduction blockade is required for antitumor activity. Due to its multitargeted effect including up- (e.g., MET) and downstream (e.g., PI3K/AKT and Ras/Raf/MEK/MAPK) signaling molecules, which are also targeted by FGF-3, HSP-90 inhibition was able to simultaneously abrogate MET, PI3K/AKT, and Ras/Raf/MEK/MAPK signaling in the resistant cells, inducing growth inhibition both in vitro and in vivo at levels largely comparable to parental cells.

Recently, it could be shown that secondary activation of EGFR signaling is another mechanism of resistance to targeted MET inhibition in MET-amplified tumor cells (30). To assess whether HSP-90 inhibition using SNX-2112 would also be effective in such tumor cells that display both MET amplification as well as secondary activation of EGFR signaling, MET-amplified NCI-H820 cells that also harbor an activating T790M EGFR mutation were used as an in vitro model (42). As shown in Supplementary Fig. S2, these cells are resistant against both EGFR and MET inhibition, whereas a greater than 50% inhibition of growth is observed on treatment with SNX-2112. Importantly, these data suggest that HSP-90 inhibition could be a promising therapeutic strategy not only to overcome different mechanisms of MET kinase inhibitor resistance (e.g., secondary activation of FGFR or EGFR signaling) but also to treat EGFR-mutant non–small-cell lung cancers that have become resistant to EGFR inhibition through secondary amplification of MET (38).

In conclusion, our study provides a strong rationale for the use of HSP-90 inhibition in tumors that have become resistant to selective tyrosine kinase inhibition and further illustrates that combinatorial signal transduction blockade offers significant advantages over highly selective kinase inhibition in oncogene-addicted malignancies.

No potential conflicts of interest were disclosed.

This work was supported by a grant from the American Society of Colon and Rectal Surgeons. T. Bachleitner-Hofmann was supported by postdoctoral research grants from the Max Kade Foundation, New York and the Austrian Surgical Society. M.R. Weiser was supported by a Career Development Award from the American Society for Clinical Oncology.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Giordano
S
,
Ponzetto
C
,
Di Renzo
MF
,
Cooper
CS
,
Comoglio
PM
. 
Tyrosine kinase receptor indistinguishable from the c-met protein
.
Nature
1989
;
339
:
155
6
.
2.
Park
M
,
Dean
M
,
Kaul
K
,
Braun
MJ
,
Gonda
MA
,
Vande Woude
G
. 
Sequence of MET protooncogene cDNA has features characteristic of the tyrosine kinase family of growth-factor receptors
.
Proc Natl Acad Sci U S A
1987
;
84
:
6379
83
.
3.
Birchmeier
C
,
Birchmeier
W
,
Gherardi
E
,
Vande Woude
GF
. 
Met, metastasis, motility and more
.
Nat Rev Mol Cell Biol
2003
;
4
:
915
25
.
4.
Francone
TD
,
Landmann
RG
,
Chen
CT
, et al
Novel xenograft model expressing human hepatocyte growth factor shows ligand-dependent growth of c-Met-expressing tumors
.
Mol Cancer Ther
2007
;
6
:
1460
6
.
5.
Lutterbach
B
,
Zeng
Q
,
Davis
LJ
, et al
Lung cancer cell lines harboring MET gene amplification are dependent on Met for growth and survival
.
Cancer Res
2007
;
67
:
2081
8
.
6.
Smolen
GA
,
Sordella
R
,
Muir
B
, et al
Amplification of MET may identify a subset of cancers with extreme sensitivity to the selective tyrosine kinase inhibitor PHA-665752
.
Proc Natl Acad Sci U S A
2006
;
103
:
2316
21
.
7.
Christensen
JG
,
Schreck
R
,
Burrows
J
, et al
A selective small molecule inhibitor of c-Met kinase inhibits c-Met-dependent phenotypes in vitro and exhibits cytoreductive antitumor activity in vivo
.
Cancer Res
2003
;
63
:
7345
55
.
8.
Christensen
JG
,
Zou
HY
,
Arango
ME
, et al
Cytoreductive antitumor activity of PF-2341066, a novel inhibitor of anaplastic lymphoma kinase and c-Met, in experimental models of anaplastic large-cell lymphoma
.
Mol Cancer Ther
2007
;
6
:
3314
22
.
9.
Zou
HY
,
Li
Q
,
Lee
JH
, et al
An orally available small-molecule inhibitor of c-Met, PF-2341066, exhibits cytoreductive antitumor efficacy through antiproliferative and antiangiogenic mechanisms
.
Cancer Res
2007
;
67
:
4408
17
.
10.
Basso
AD
,
Solit
DB
,
Chiosis
G
,
Giri
B
,
Tsichlis
P
,
Rosen
N
. 
Akt forms an intracellular complex with heat shock protein 90 (Hsp90) and Cdc37 and is destabilized by inhibitors of Hsp90 function
.
J Biol Chem
2002
;
277
:
39858
66
.
11.
Basso
AD
,
Solit
DB
,
Munster
PN
,
Rosen
N
. 
Ansamycin antibiotics inhibit Akt activation and cyclin D expression in breast cancer cells that overexpress HER2
.
Oncogene
2002
;
21
:
1159
66
.
12.
Sawai
A
,
Chandarlapaty
S
,
Greulich
H
, et al
Inhibition of Hsp90 down-regulates mutant epidermal growth factor receptor (EGFR) expression and sensitizes EGFR mutant tumors to paclitaxel
.
Cancer Res
2008
;
68
:
589
96
.
13.
Schulte
TW
,
Blagosklonny
MV
,
Romanova
L
, et al
Destabilization of Raf-1 by geldanamycin leads to disruption of the Raf-1-MEK-mitogen-activated protein kinase signalling pathway
.
Mol Cell Biol
1996
;
16
:
5839
45
.
14.
Shimamura
T
,
Li
D
,
Ji
H
, et al
Hsp90 inhibition suppresses mutant EGFR-T790M signaling and overcomes kinase inhibitor resistance
.
Cancer Res
2008
;
68
:
5827
38
.
15.
Shimamura
T
,
Lowell
AM
,
Engelman
JA
,
Shapiro
GI
. 
Epidermal growth factor receptors harboring kinase domain mutations associate with the heat shock protein 90 chaperone and are destabilized following exposure to geldanamycins
.
Cancer Res
2005
;
65
:
6401
8
.
16.
Solit
DB
,
Zheng
FF
,
Drobnjak
M
, et al
17-Allylamino-17-demethoxygeldanamycin induces the degradation of androgen receptor and HER-2/neu and inhibits the growth of prostate cancer xenografts
.
Clin Cancer Res
2002
;
8
:
986
93
.
17.
Whitesell
L
,
Mimnaugh
EG
,
De Costa
B
,
Myers
CE
,
Neckers
LM
. 
Inhibition of heat shock protein HSP90-pp60v-src heteroprotein complex formation by benzoquinone ansamycins: essential role for stress proteins in oncogenic transformation
.
Proc Natl Acad Sci U S A
1994
;
91
:
8324
8
.
18.
Banerji
U
. 
Heat shock protein 90 as a drug target: some like it hot
.
Clin Cancer Res
2009
;
15
:
9
14
.
19.
Bachleitner-Hofmann
T
,
Sun
MY
,
Chen
CT
, et al
HER kinase activation confers resistance to MET tyrosine kinase inhibition in MET oncogene-addicted gastric cancer cells
.
Mol Cancer Ther
2008
;
7
:
3499
508
.
20.
Koga
F
,
Tsutsumi
S
,
Neckers
LM
. 
Low dose geldanamycin inhibits hepatocyte growth factor and hypoxia-stimulated invasion of cancer cells
.
Cell Cycle
2007
;
6
:
1393
402
.
21.
Shen
Y
,
Xie
Q
,
Norberg
M
,
Sausville
E
,
Woude
GV
,
Wenkert
D
. 
Geldanamycin derivative inhibition of HGF/SF-mediated Met tyrosine kinase receptor-dependent urokinase-plasminogen activation
.
Bioorg Med Chem
2005
;
13
:
4960
71
.
22.
Webb
CP
,
Hose
CD
,
Koochekpour
S
, et al
The geldanamycins are potent inhibitors of the hepatocyte growth factor/scatter factor-met-urokinase plasminogen activator-plasmin proteolytic network
.
Cancer Res
2000
;
60
:
342
9
.
23.
Xie
Q
,
Gao
CF
,
Shinomiya
N
, et al
Geldanamycins exquisitely inhibit HGF/SF-mediated tumor cell invasion
.
Oncogene
2005
;
24
:
3697
707
.
24.
Rice
JW
,
Veal
JM
,
Barabasz
A
, et al
Targeting of multiple signaling pathways by the Hsp90 inhibitor SNX-2112 in EGFR resistance models as a single agent or in combination with erlotinib
.
Oncol Res
2009
;
18
:
229
42
.
25.
Chandarlapaty
S
,
Sawai
A
,
Ye
Q
, et al
SNX2112, a synthetic heat shock protein 90 inhibitor, has potent antitumor activity against HER kinase-dependent cancers
.
Clin Cancer Res
2008
;
14
:
240
8
.
26.
Jin
L
,
Xiao
CL
,
Lu
CH
, et al
Transcriptomic and proteomic approach to studying SNX-2112-induced K562 cells apoptosis and anti-leukemia activity in K562-NOD/SCID mice
.
FEBS Lett
2009
;
583
:
1859
66
.
27.
Okawa
Y
,
Hideshima
T
,
Steed
P
, et al
SNX-2112, a selective Hsp90 inhibitor, potently inhibits tumor cell growth, angiogenesis, and osteoclastogenesis in multiple myeloma and other hematologic tumors by abrogating signaling via Akt and ERK
.
Blood
2009
;
113
:
846
55
.
28.
Kammula
US
,
Kuntz
EJ
,
Francone
TD
, et al
Molecular co-expression of the c-Met oncogene and hepatocyte growth factor in primary colon cancer predicts tumor stage and clinical outcome
.
Cancer Lett
2007
;
248
:
219
28
.
29.
Ornitz
DM
,
Xu
J
,
Colvin
JS
, et al
Receptor specificity of the fibroblast growth factor family
.
J Biol Chem
1996
;
271
:
15292
7
.
30.
McDermott
U
,
Pusapati
RV
,
Christensen
JG
,
Gray
NS
,
Settleman
J
. 
Acquired resistance of non-small cell lung cancer cells to MET kinase inhibition is mediated by a switch to epidermal growth factor receptor dependency
.
Cancer Res
70
:
1625
34
.
31.
Jhawer
M
,
Kindler
HL
,
Wainberg
Z
, et al
Assessment of two dosing schedules of GSK1363089 (GSK089), a dual MET/VEGFR2 inhibitor, in metastatic gastric cancer (GC): interim results of a multicenter phase II study [Meeting Abstracts]
.
J Clin Oncol
2009
;
27
:
4502
.
32.
Kwak
EL
,
Camidge
DR
,
Clark
J
, et al
Clinical activity observed in a phase I dose escalation trial of an oral c-met and ALK inhibitor, PF-02341066 [Meeting Abstracts]
.
J Clin Oncol
2009
;
27
:
3509
.
33.
Galdemard
C
,
Brison
O
,
Lavialle
C
. 
The proto-oncogene FGF-3 is constitutively expressed in tumorigenic, but not in non-tumorigenic, clones of a human colon carcinoma cell line
.
Oncogene
1995
;
10
:
2331
42
.
34.
Galdemard
C
,
Yamagata
H
,
Brison
O
,
Lavialle
C
. 
Regulation of FGF-3 gene expression in tumorigenic and non-tumorigenic clones of a human colon carcinoma cell line
.
J Biol Chem
2000
;
275
:
17364
73
.
35.
Hajitou
A
,
Baramova
EN
,
Bajou
K
, et al
FGF-3 and FGF-4 elicit distinct oncogenic properties in mouse mammary myoepithelial cells
.
Oncogene
1998
;
17
:
2059
71
.
36.
Hajitou
A
,
Calberg-Bacq
CM
. 
Fibroblast growth factor 3 is tumorigenic for mouse mammary cells orthotopically implanted in nude mice
.
Int J Cancer
1995
;
63
:
702
9
.
37.
Hajitou
A
,
Deroanne
C
,
Noel
A
, et al
Progression in MCF-7 breast cancer cell tumorigenicity: compared effect of FGF-3 and FGF-4
.
Breast Cancer Res Treat
2000
;
60
:
15
28
.
38.
Engelman
JA
,
Zejnullahu
K
,
Mitsudomi
T
, et al
MET amplification leads to gefitinib resistance in lung cancer by activating ERBB3 signaling
.
Science
2007
;
316
:
1039
43
.
39.
Guix
M
,
Faber
AC
,
Wang
SE
, et al
Acquired resistance to EGFR tyrosine kinase inhibitors in cancer cells is mediated by loss of IGF-binding proteins
.
J Clin Invest
2008
;
118
:
2609
19
.
40.
Marek
L
,
Ware
KE
,
Fritzsche
A
, et al
Fibroblast growth factor (FGF) and FGF receptor-mediated autocrine signaling in non-small-cell lung cancer cells
.
Mol Pharmacol
2009
;
75
:
196
207
.
41.
Workman
P
,
Burrows
F
,
Neckers
L
,
Rosen
N
. 
Drugging the cancer chaperone HSP90: combinatorial therapeutic exploitation of oncogene addiction and tumor stress
.
Ann N Y Acad Sci
2007
;
1113
:
202
16
.
42.
Bean
J
,
Brennan
C
,
Shih
JY
, et al
MET amplification occurs with or without T790M mutations in EGFR mutant lung tumors with acquired resistance to gefitinib or erlotinib
.
Proc Natl Acad Sci U S A
2007
;
104
:
20932
7
.